Alibaba officially released the new large model Qwen3.8-Max. According to the official introduction of the Qwen large model, the new model has achieved comprehensive upgrades in core capabilities such as programming, real office work, long-term task execution, and multimodal agents, further enhancing the ability to autonomously execute and deliver complex tasks. The model weights are planned to be officially open-sourced next week.
The official stated that Qwen3.8-Max can complete complex and open-ended tasks with minimal human intervention, achieving end-to-end execution from goal understanding, task planning, to final delivery, with enhanced application capabilities of agents in real production environments. This upgrade covers not only code generation and software development scenarios but also further optimizes office automation, multi-step reasoning, and multimodal collaborative processing capabilities, enabling the model to handle more complex practical workflows.
The upcoming open-source of model weights means developers and enterprises can perform secondary development, model fine-tuning, and intelligent agent application building based on Qwen3.8-Max, further promoting the development of the open-source large model ecosystem.
As industry competition gradually shifts from basic model performance to agent capabilities and the efficiency of complex task execution, the release of Qwen3.8-Max further reflects the development trend of large models towards high autonomy, end-to-end delivery, and real business scenario deployment.
